Topics

Are your transistors growing whiskers in their old age?

Dave Seiter
 

So does silver.
-Dave

On Tuesday, March 3, 2020, 07:58:34 AM PST, Brad Thompson <brad.thompsonaa1ip@...> wrote:

Geoffrey Thomas wrote on 3/3/2020 9:53 AM:

The AF117 transistors were exhibiting this problem in the late 1970's,
tin whiskers is why we had lead in solder, of course.
Hello--

Zinc also grows whiskers:

https://www.cableorganizer.com/learning-center/articles/zinc-whiskers.html

73--

Brad  AA1IP

stevenhorii
 

My interest in restoration of electronics stuff is mostly frequency/time
standards. The ones I work on are from the ‘60s and later. Most have
discrete components. These are mostly ovenized quartz oscillators. The ones
I have had most experience with are the Tracor/Sulzer series. Some of these
I have had in operation for over 20 years. When I called Tracor, they still
were able to get me copies of the manuals, though these were not cheap. I
told them I was still using 30-40 year old oscillators and they were not
surprised.

Failures have been more common in the power supplies - the design of the
power supply includes a NiCd battery pack that is float charged. If a cell
goes bad, the supply will switch into a high charge rate mode. This
eventually results in the transformer (sealed in a metal can) overheating.
I have had two supplies go bad and when I opened the power supply, found
that the insides of the case encasing the supply was partially filled with
the tar-like dielectric (?heat sinking material) from the transformer. The
transformer case had split along the header and cover. Surprisingly, the
other parts were OK, though a resistor showed evidence of having run hot.
The transistors are silicon ones and I have had a couple go bad. The
replacements were the plastic-cased ones that were the same number but in
the plastic cases. I wonder if that plastic inhibits the crystal growth.
Cleaning that mess up was not worth it. Since I was running several of the
standards, I just used a single regulated supply and didn’t worry about a
battery backup (though I did, for awhile, use a PC battery backup).

Given this discussion and the age of the transistors, I would not be
surprised if the transistor failure mode was crystal growth between the
small leads off the die.

This PDF of a presentation on reliability by Chandra Gupta at the IEEE
Boston meeting on reliability discusses both whisker growth and something
called the Kirkendall effect. This is when voids open in the small gold
wires bonded to the pads on and IC chip. It sounds like there is a
long-term lifecycle potential problem with ICs as well.

http://ewh.ieee.org/r1/boston/rl/files/boston_rs_meeting_sep17.pdf

The stuff on whiskers and the Kirkendall effect is nearer the end of the
presentation.

Steve

On Tue, Mar 3, 2020 at 11:43 greenboxmaven via Groups.Io <ka2ivy=
verizon.net@groups.io> wrote:

I also enjoy restoring older complex portable radios. The Mullard
transistors used in early solid state Eddystone receivers are famous for
failures. Some people disconnect the case/ground lead, others connect
the E-B-C leads together and apply a condenser charged to about 100
volts to the leads and the case to fry the whiskers. I felt there was
nothing to lose for a more radical repair. These transistors had fairly
large cases and were obviously soldered together. I unsoldered the top
of the case from the header and swept the leads with an artist's brush,
taking care to not touch the crystal, which was embedded in silicon
grease. I then cleaned the case out with the brush, and reassembled the
transistor. My success was 100% in eliminating the shorts or leakage,
and about 90% overall. The receivers are working fine today. As early
solid state equipment ages, this problem will become worse. To my
knowledge, there has not been any problem like this with molded plastic
transistors. It does occur in ICs because there is space around the
silicon slab. I wonder if any of the designers back then could have
imagined the equipment being restored and used 50 or more years later.

Bruce Gentry, KA2IVY


On 3/3/20 10:58 AM, Brad Thompson wrote:
Geoffrey Thomas wrote on 3/3/2020 9:53 AM:

The AF117 transistors were exhibiting this problem in the late
1970's, tin whiskers is why we had lead in solder, of course.






Dave Wise
 

My 321A also has whiskery transistors. Rather than tapping, I vaporized the whiskers by applying a few volts between the case and whichever element had shorted to the case. That was years ago and any still growing haven't shorted yet.

Some parts stayed dead; the zap must have burned out a junction. Not having any 2N2207's in stock at the time, I improvised with silicon. This got very interesting in the timebase, as the idle operating point depends on a closed-loop chain of Vbe's and diode drops.

Dave Wise
________________________________________
From: TekScopes@groups.io <TekScopes@groups.io> on behalf of John Williams via Groups.Io <books4you=telus.net@groups.io>
Sent: Monday, March 02, 2020 8:00 PM
To: TekScopes@groups.io
Subject: [TekScopes] Are your transistors growing whiskers in their old age?

I suspect that many of you already know about this problem, but it is new to me. It relates to a number of 1960s era transistors causing failures in transistorized radios. The types of transistors described to have this behaviour are also used in early Tektronix equipment including the 321 and 321A oscilloscopes. According to the UK radio restorers people and investigated by NASA, tin “whiskers “ develop over time between the inner transistor case and connections inside. This causes hard to fix problems in the affected equipment. One temporary fix is actually tapping the transistor itself. I have a number of these little scopes that I parted out because I could not find the problem. Last night I actually fixed a 321 by tapping some of the transistors. These scopes have quite a few of the transistors susceptible to this problem, notably 2N2207 and OC170.

Information on this can be found by reading and following links within the following documents:

https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html

http://www.vintage-radio.com/repair-restore-information/transistor_transistor-faults.html

https://nepp.nasa.gov/whisker/

There is a lot more but this is a good start. I now must go tap some more transistors.

Geoffrey Thomas
 

That's why it's used in galvanising to protect against rust, any holes and the zinc spreads itself out to seal the hole. Rustproofing fails when the zinc is depleted.

Geoff.

On 03/03/2020 15:58, Brad Thompson wrote:
Geoffrey Thomas wrote on 3/3/2020 9:53 AM:

The AF117 transistors were exhibiting this problem in the late 1970's, tin whiskers is why we had lead in solder, of course.
Hello--
Zinc also grows whiskers:
https://www.cableorganizer.com/learning-center/articles/zinc-whiskers.html
73--
Brad  AA1IP

greenboxmaven
 

I also enjoy restoring older complex portable radios. The Mullard transistors used in early solid state Eddystone receivers are famous for failures. Some people disconnect the case/ground lead, others connect the E-B-C leads together and apply a condenser charged to about 100 volts to the leads and the case to fry the whiskers. I felt there was nothing to lose for a more radical repair. These transistors had fairly large cases and were obviously soldered together. I unsoldered the top of the case from the header and swept the leads with an artist's brush, taking care to not touch the crystal, which was embedded in silicon grease. I then cleaned the case out with the brush, and reassembled the transistor. My success was 100% in eliminating the shorts or leakage, and about 90% overall. The receivers are working fine today. As early solid state equipment ages, this problem will become worse. To my knowledge, there has not been any problem like this with molded plastic transistors. It does occur in ICs because there is space around the silicon slab. I wonder if any of the designers back then could have imagined the equipment being restored and used 50 or more years later.

Bruce Gentry, KA2IVY

On 3/3/20 10:58 AM, Brad Thompson wrote:
Geoffrey Thomas wrote on 3/3/2020 9:53 AM:

The AF117 transistors were exhibiting this problem in the late 1970's, tin whiskers is why we had lead in solder, of course.


Brad Thompson
 

Geoffrey Thomas wrote on 3/3/2020 9:53 AM:

The AF117 transistors were exhibiting this problem in the late 1970's, tin whiskers is why we had lead in solder, of course.
Hello--

Zinc also grows whiskers:

https://www.cableorganizer.com/learning-center/articles/zinc-whiskers.html

73--

Brad  AA1IP

Geoffrey Thomas
 

The AF117 transistors were exhibiting this problem in the late 1970's, tin whiskers is why we had lead in solder, of course.

Geoff.

On 03/03/2020 14:44, bill koski via Groups.Io wrote:
I restore vintage radios and have had a couple early transistor radios that I had to replace a germanium transistor that I suspect failed from whiskers. The first transistor radio to be mass marketed was the Regency TR1. From what I read the first run of those radios were recalled because of early mortality of the transistors. Finding one of those first run radios is rare and a working one that is all original is even more rare.

bill koski
 

I restore vintage radios and have had a couple early transistor radios that I had to replace a germanium transistor that I suspect failed from whiskers. The first transistor radio to be mass marketed was the Regency TR1. From what I read the first run of those radios were recalled because of early mortality of the transistors. Finding one of those first run radios is rare and a working one that is all original is even more rare.

David Holland
 

I've not seen it in transistors, but I have seen it in a potentiometer
once. (B&W TV contrast control IIRC). High voltage seems to make it more
likely - also IIRC :-)

David

Sent via mobile annoyance thingy, please pardon any typos.

On Mon, Mar 2, 2020, 11:00 PM John Williams <books4you@...> wrote:

I suspect that many of you already know about this problem, but it is new
to me. It relates to a number of 1960s era transistors causing failures in
transistorized radios. The types of transistors described to have this
behaviour are also used in early Tektronix equipment including the 321 and
321A oscilloscopes. According to the UK radio restorers people and
investigated by NASA, tin “whiskers “ develop over time between the inner
transistor case and connections inside. This causes hard to fix problems in
the affected equipment. One temporary fix is actually tapping the
transistor itself. I have a number of these little scopes that I parted out
because I could not find the problem. Last night I actually fixed a 321 by
tapping some of the transistors. These scopes have quite a few of the
transistors susceptible to this problem, notably 2N2207 and OC170.

Information on this can be found by reading and following links within the
following documents:

https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html


http://www.vintage-radio.com/repair-restore-information/transistor_transistor-faults.html

https://nepp.nasa.gov/whisker/

There is a lot more but this is a good start. I now must go tap some more
transistors.






John Williams
 

I suspect that many of you already know about this problem, but it is new to me. It relates to a number of 1960s era transistors causing failures in transistorized radios. The types of transistors described to have this behaviour are also used in early Tektronix equipment including the 321 and 321A oscilloscopes. According to the UK radio restorers people and investigated by NASA, tin “whiskers “ develop over time between the inner transistor case and connections inside. This causes hard to fix problems in the affected equipment. One temporary fix is actually tapping the transistor itself. I have a number of these little scopes that I parted out because I could not find the problem. Last night I actually fixed a 321 by tapping some of the transistors. These scopes have quite a few of the transistors susceptible to this problem, notably 2N2207 and OC170.

Information on this can be found by reading and following links within the following documents:

https://nepp.nasa.gov/whisker/anecdote/af114-transistor/index.html

http://www.vintage-radio.com/repair-restore-information/transistor_transistor-faults.html

https://nepp.nasa.gov/whisker/

There is a lot more but this is a good start. I now must go tap some more transistors.